《The Global City&畢馬威：2024專業金融服務：人工智能和勞動力的未來（英文版）(52頁）.pdf》由會員分享，可在線閱讀，更多相關《The Global City&畢馬威：2024專業金融服務：人工智能和勞動力的未來（英文版）(52頁）.pdf（52頁珍藏版）》請在三個皮匠報告上搜索。

1、Financial&professional services:The future of AI&the workforceCity of London Corporation&KPMG June 2024In association withForewordIn recent months,the sudden and very public exposure of generative Artificial Intelligence(AI)has created immense debate about the broader good,and potential risks,that A

2、I and Machine Learning can bring to society and the economy.From the overall questions of ethics and governance to sector-specific applications for innovative new use cases,the UK stands to be a global leader in this transformational new technology.Yet,as we stand on the precipice of a Fourth Indust

3、rial Revolution,marked by the rapid integration of AI and automation into our businesses,it is imperative that we pause to assess the implications of these transformative technologies on our workforce.It is always our shared aim to ensure that our Financial and Professional Services(FPS)sector be th

4、e best-in-class,driving UK plc,creating the wealth and jobs that support the economy and public services.That is why in our Vision for Economic Growth,the City of London Corporation highlighted the need to be a digital-first economy which leads on AI and machine learning.It is why we are doing lots

5、of work on AI to understand the opportunities,wider economic and societal impacts,regulatory and ethical considerations.1 Our UK financial and professional services industry is world-leading and globally respected.The UK tech ecosystem is widely considered one of the best in the world and this keeps

6、 global talent coming to innovative UK companies.The UK fintech sector attracts more than three times the investment in France,Germany,Hong Kong,Japan and Singapore combined,and globally ranking second only to the US2.According to our recent report AI-Accelerating Innovation,the FPS sector is a lead

7、er in AI adoption and the UK is well-positioned to be a global leader in AI,with the third highest AI private capital behind the US and China3.However,there is a national tech skills gap.The Financial Services Skills Commissions recent Skills for the future of Financial Services 2024 notes that mach

8、ine learning/AI has the biggest supply-demand gap for its members4.At the same time,recent Government reports highlight high levels of exposure to the FPS sector Finance and Insurance being the most exposed industries5.Jobs will change,and skillsets will need to adapt.To date providing that granular

9、 understanding of how AI and automation will impact the FPS sector by subsectors,business functions and even tasks has been wanting.Without a better understanding of how it is already impacting the sector,we stand little chance of harnessing its great potential to improve our productivity,our innova

10、tion,and critically,our people.That is why we are pleased to be publishing this report.With a more detailed look at current and potential future use cases for our sector specifically,we can begin to understand which tasks could increasingly be automated,and in turn spot where skillsets and opportuni

11、ties for the years to come will be realised.By undertaking a comprehensive assessment of AI impacts on our workforce,we reaffirm our commitment to building a future where innovation serves as a catalyst for prosperity,equality,and inclusive growth.I recommend this report to all those who share in ou

12、r vision for a UK workforce that is not only prepared for the challenges of tomorrow but poised to seize the opportunities of today.Chris Hayward Policy Chairman of the City of London Corporation Financial&Professional Services:The Future of AI&the workforceIntroduction2The AI journey so far As a yo

13、ung Computer Science undergraduate in the 80s,the terms machine learning and AI,as well as robotics,were already in regular use,and we were taught the basic concepts.However,the actual application of these to anything affecting everyday life was still very much a part of the science fiction films en

14、joyed at the campus cinema.After all,this was still the time of bleepers,PCs with fixed keyboards and green screens,and the BBC computer feels like light years away from today.Behind the scenes,academics and research departments have continued to develop in the AI sphere,moving at pace,but without s

15、ignificant impact on the general public or workplace.Generative AI has now exploded into our business sphere,promising substantial benefits to efficiency and cost,removing those administrative tasks we are regularly burdened with,and freeing us up to think innovatively and grow faster.It is an excit

16、ing opportunity and set to speed up the technology revolution.Many CEOs have welcomed this new technology,encouraging organisations to embrace it and apply it wherever possible,and at speed sometimes without fully understanding the full impact,or the associated risks.Dont get me wrong there is much

17、to be gained by embracing AI,but we need to do so in a measured way,ensuring we are enabling employees to understand and exploit the new technology without introducing risk to our businesses.30 years ago,outsourcing was the nirvana for processing efficiency and it seemed we would all work a 3 day we

18、ek,retire at 50,and operate in a paperless environment.With a strategic approach to the adoption of AI,hopefully we can actually make some of these things a reality!Mel Newton,KPMG Partner,FS People ConsultingThe evolution of AIArtificial intelligence 1956 The field of computer science that seeks to

19、 create intelligent machines that can replicate or exceed human intelligence1997Machine learningSubset of AI that enables machines to learn from existing data and improve upon that data to make decisions or predications2017Deep learning A machine learning technique in which layers of neural networks

20、 are used to process data and make decisions (these are called transformers)2021Generative AICreate new written,visual,and auditory content given prompts or existing data.For example,Large Language Models like ChatGPT3Financial&Professional Services:The Future of AI&the WorkforceIntroductionContents

21、5Section 1:Introduction and key findings10Section 2:The projected impact of AI on the UKs financial and professional services workforce 32Section 3:What employers should consider to maximise the benefits of using AI in their organisations41AppendixIntroductionAI has become an important element of th

22、e workforce transformation journey,creating endless possibilities to change the way business is conducted.Whist it continues to be the headline topic for many,its impact on the workforce continues to be speculative.The purpose of this report is to take a closer look at the expected impact of AI on t

23、he FPS as a whole,as well as sub-sectors inclusive of retail banking,fintech,challenger banks,investment banks,asset managers,insurance and legal services.With this in mind,we discuss what employers should consider to maximise the benefits of AI within their organisation.Although AI is not new to th

24、e FPS sector,the recent explosion of generative AI has,for the first time,seen knowledge work(work requiring an element of thinking)automated.Until recently,this type of work was thought to reside exclusively in the human domain.GenAI will transform nations,sectors,companies and individuals at pace.

25、This report by the City of London Corporation and KPMG is leading the debate.Considering how to capture the value of AI,manage its risks,partner for success and still hold trust with our people and our teams is the biggest challenge of the 2020s.”“David RowlandsKPMG Partner,Global Head of AI5Financi

26、al&Professional Services:The Future of AI&the WorkforceSection 1:Introduction&key findingsThere is,understandably,a degree of unease about the potential impact of generative AI.Yet in the course of the last thirty years,comparable levels of disruption have been wrought by technological advancements

27、in computing power and most importantly the internet.While these changes have reduced the demand for some skillsets(such as bookkeeping and certain secretarial services)they have created vast and important efficiencies and new roles.The impact of AI on the FPS sector is particularly exciting as it p

28、resents an opportunity to address.many of the workforce challenges the growing sector is facing.Research by the ONS shows that the number of unfilled vacancies in the FPS sector is growing6 and the demand for technical and behavioural competencies outstrips their supply to the tune of 20 percent and

29、 25 percent,respectively.7 In addition,64 percent of employees say they lack the energy needed to do their jobs8,and 70 percent agreed that they would delegate as much work as possible to AI to lessen their load.9 Overall,the FPS sector is struggling to attract and retain the talent needed to delive

30、r,and colleagues within the industry report increasingly high workloads with long-term sickness and stress-leave at their highest rates ever10.This points to two key considerations.The first is that the skills required by organisations are changing the World Economic Forum(WEF)has projected that 44

31、percent of workers core skills will change within five years11 and FPS firms should be supporting colleague skill development in this vein.The second is that the workforce is overstretched.Productivity in the sector may be on the rise,but there remains a significant opportunity to lighten the burden

32、 placed on colleagues and leverage new developments in AI to enhance the way work is done,freeing up time for higher-value activities and driving growth.The advancements of AI create an exciting new stage in the digital revolution.It is here to be harnessed and will accelerate the pace of workforce

33、transformation bringing many benefits to employers,employees,and customers.While the specifics of this future workforce are unclear,the use of AI will be extensive,with workers focused more on collaboration and adding value.Firms in the FPS sector should be focused on what they can and need to be do

34、ing in order to optimise the adoption of AI or the benefit of the workforce as well as shareholders.6Financial&Professional Services:The Future of AI&the WorkforceSection 1:Introduction&key findingsKey findings1Adoption of AI within the FPS sector in the UK has been in line with the maturity and cau

35、tion of a highly regulated sector,with demand for AI skills increasing from a relatively low base.As organisations develop infra-structure&controls that can responsibly scale Generative AI solutions,and the regulatory landscape becomes clearer and better understood,adoption is set to increase substa

36、ntially.Adoption2As Generative AI is data-driven with a degree of autonomy,it poses serious risk if not managed correctly.Complexity of legacy systems and data quality,cost-consciousness,ethical concerns such as data privacy and explainability,the regulatory environment,operational risks and halluci

37、nations,and the potential impact on employee engagement and culture all contribute to a complex array of risks to be considered and managed.Risk3OpportunitiesProjections indicate that in the short term(one year),the FPS sector could experience a 12 percent uplift in productivity by adopting availabl

38、e and relatively easy-to-implement AI solutions.Over five years,provided AI is fully integrated and put to maximum advantage,the gains could be as high as 50 percent.The greatest long-term opportunity for AI powered growth in the sector resides in hyper-personalised products and expansion into new m

39、arkets and demographics.4SkillsWhile overall the number of FPS job postings that include AI skills is relatively low,we are seeing rapid growth in the demand for certain skills(e.g.150 x for generative&conversational AI)as well as growth in demand for skills required to support AI work.The current s

40、upply of relevant skills in the UK workforce does not meet current demand levels,even less so the anticipated requirement of the near future.Research suggests that delivering training on using AI generates an immediate observable benefit on productivity;its highly commercial and low risk for compani

41、es to mobilise quickly in AI training provision.5TalentWhile AI may be able to augment and replace many tasks,it will also create new jobs for people requiring new skills.With the cost of recruitment in the FPS sector so high and the reality that many of the AI and AI-adjacent skills of the future a

42、re still relatively new and not yet prevalent in the labour market,employers should consider bridging the AI knowledge gap via training and development rather than by looking to recruit AI talent in the scarce and highly competitive marketplace.6PolicyWhilst the regulatory landscape is developing,it

43、s important for companies to show leadership in the development of AI governance and contribute to the dialogue.In relation to the national market for talent,the UK is adopting special visas in the short term to supply AI skills from aboard and investing in education provision and apprenticeships to

44、 meet longer term demand domestically.It is critical for public,private and educational institutions to work together so the UK workforce can capitalise on this historic opportunity.7Financial&Professional Services:The Future of AI&the WorkforceSection 1:Introduction&key findingsKey findings:The num

45、bersProjected FPS sector productivity boost by 2030 50%Our research suggests the FPS sector could see an increase in productivity of 12 percent over the next year if firms implement and adopt“general augmentation”AI product firmwide,and 50 percent over the next five years if all opportunities for AI

46、 enhancement of current business operations are taken(see Appendix for methodology).Productivity boost for support functions over 5 years58%Support functions are projected to have the largest capacity to leverage AI to improve productivity.The ongoing importance of relationships and customer interac

47、tion in Revenue functions underpins lower projected levels of workforce impact by AI.Control functions productivity boost is limited by the need of draft,interpret,and enforce policies and controls.Productivity boost year 1Productivity boost years 2-538%12%SupportControlRevenueProductivity boost yea

48、r 1Productivity boost years 2-545%13%12%11%37%33%(5 year total:50%)(5 year total:58%)(5 year total:49%)(5 year total:44%)8Financial&Professional Services:The Future of AI&the WorkforceSection 1:Introduction&key findingsProjected AI driven value-add to FPS sector by 2030 35bnBased on the size of the

49、sector,AI adoption can add significant value through productivity gains,enabling growth in new markets,and product development.Value will be realised across the FPS sector,with retail banking,insurance and legal services benefitting the most based on projected AI adoption.150 xIncrease in demand for

50、 generative and conversational AI skills since 2021 While overall the number of job postings across the FPS sector that include AI skills is relatively low,we are seeing rapid growth in the demand for certain AI skills(e.g.150 x for generative&conversational AI)as well as growth in demand for skills

51、 required to support AI work.Looking across FPS sectors,fintech and challenger banks have the highest demand for AI skills,whilst the demand for AI skills is growing the fastest within legal services and retail banking.Overall the demand is relatively low,ranging from 1.2-5.4 percent of job postings

52、 including AI skills in the last 12 months.FPS Sector increasing demand in AI skills 2.6bn2.4bn1.3bn9.0bn9.3bn9.9bn0.6bnAsset ManagementChallenger BanksFintechInsuranceInvestment BankingLegal ServicesRetail BankingAnalytics&Data ScienceGenerative&Conversational AIMetricInsuranceAsset managementFinte

53、chLegal servicesChallenger banksInvestment bankingRetail banking%Job posts in the last 12 months include AI skills2.6%4.7%5.4%1.2%4.9%2.7%4.5%Increase in demand for AI skills in the last three years90%52%40%345%196%76%236%1.6x1.8x2x3x15x150 xFPS sector increasing demand in AI skillsFS sector total A

54、I value-addSupporting Transferable SkillsAI&ML ConceptAI&ML TechniquesGovernance&Ops9Financial&Professional Services:The Future of AI&the WorkforceSection 1:Introduction&key findingsSection 2 The projected impact of AI on the UKs financial and professional services workforce ApproachThe FPS services

55、 sector incorporates a wide range of businesses and their activities which are crucial to the functioning of the modern economy.For the purposes of this report,we have focused on the following areas:FPS sectors reviewed FPS business areas reviewed:Revenue functionsFunctions responsible for generatin

56、g income and driving business growthControl functionsFunctions that oversee business operations comply with legal and regulatory guidance,as well as the organisations own risk appetiteSupport functionsRoles within functions that furnish services to enable effective operations(e.g.IT,Operations,Human

57、 Resources,Finance)Retail banking Insurance Legal servicesFintechChallenger banksInvestment banking Asset management 11Financial&Professional Services:The Future of AI&the WorkforceSection 2:The projected impact of AI on the UKs financial and professional services workforceAsset management:Impact of

58、 AIAsset management has embarked on its AI journey;even so,greater opportunities exist to bolster performance and employee experience.Though many firms have reaped the fruits of using AI in positions such as those of data analyst,portfolio manager,and market research specialist,they have scarcely sc

59、ratched the surface of what will come to be an all-embracing reinvention of job roles.Its application has thus far been characterised by the monitoring of market trends,the culling of information,and the generation of content.AI skills across the asset management workforce of job posts in the last 1

60、2 months include AI skills+52%increased demand for AI skills in the last three years AI adoption is starting to grow but AI skills within the workforce are lagging.4.7%The figure below shows the spread of AI skills being seen in job posts in the last 12 months.Please see appendix for methodology.Sec

61、tion 2:The projected impact of AI on the UKs financial and professional services workforceGen AI&ChatbotsAI&ML TechniquesAI&ML Governance&OperationsAI&ML Core74382125034Asset managementVery fast growth.Either 150%OR any skill with a 0 baseline in 3YA data(i.e.100%growth)Fast growth.Between 50%and 14

62、9.9%changeLower growth.Below 49.9%changeGraph 1:AI skills in asset managementAI skill groupsJob posts in last 12-months03006009001200150012Financial&Professional Services:The Future of AI&the WorkforceChallengesSkillsThe shortage in todays asset management workforce of the skills needed to identify

63、and effectively implement use cases for AI tools is inhibiting progress.Data qualityAssuring data quality remains a challenge for asset managers.Many firms rely on third-party market data;their own must be improved before the benefits of AI can be unlocked.RegulationThe uncertainty surrounding AI re

64、gulation is a concern for leaders in this space,as limitations on accessible guidance cast doubt over the longevity of AI investment cases.Section 2:The projected impact of AI on the UKs financial and professional services workforce13Financial&Professional Services:The Future of AI&the WorkforceSect

65、or use cases Customer-centric AI The implementation of a customer-focused AI solution such as Microsoft Copilot or Salesforce could allow for more personalised client experiences,meanwhile freeing the value-add capacity of client relationship managers in their every touchpoint.The deployment of AI t

66、o track the lineage of client conversations alongside the ability of predictive and scenario analysis to scope trends and simulates various interactions can help better prepare and equip relationship managers with necessary insights.AI-powered chatbots and virtual assistants are capable of providing

67、 service to clients autonomously,enhancing the latters overall experience and easing the strain on relationship managers.Productivity and advisory capacity AI is adept at automating routine tasks such as data entry,reconciliation,and report generation,freeing up time for asset managers to focus on m

68、ore strategic activities.There will be a further movement towards using Large Language Models(LLMs)for the digestion of vast amounts of data from various sources to identify investment opportunities,conduct market research,and generate insights that will inform investment decisions more rapidly.Insu

69、rance:Impact of AIThe implementation of AI in insurance represents a significant technological advancement in the industrys evolution.Traditionally,the insurance sector has been cautious in adopting major digital innovations.Key areas such as underwriting,actuarial work,and claims handling are exper

70、iencing notable changes,with clear applications for enhancing customer experience.Existing machine learning models,which have been developing within the industry for several years,stand to benefit from the integration of AI.Despite the considerable interest in AI,insurance companies typically evalua

71、te investments individually,focusing on those that provide tangible value to customers.AI skills across the insurance workforce+90%increased demand for AI skills in the last three years Insurance firms are some way from fully capitalising on the benefits of an AI-skilled workforce.2.6%The figure bel

72、ow shows the spread of AI skills being seen in job posts in the last 12 months.Please see appendix for methodology.of job posts in the last 12 months include AI skillsInsuranceSection 2:The projected impact of AI on the UKs financial and professional services workforceGen AI&ChatbotsAI&ML Techniques

73、AI&ML Core6012438515Graph 2:AI skills in insuranceAI skill groupsJob posts in last 12-monthsVery fast growth.Either 150%OR any skill with a 0 baseline in 3YA data(i.e.100%growth)Fast growth.Between 50%and 149.9%changeLower growth.Below 49.9%changeAI&ML Governance&Operations03006009001200150014Financ

74、ial&Professional Services:The Future of AI&the WorkforceChallengesLegacy systemsThe data required to power AI is stored in many legacy systems,hamstringing implementation.Employee engagementEmployees often require greater clarity on the strategic vision and purpose of AI transformation,expressing a

75、desire for more involvement in understanding what it means for them.Return On investmentIn a cost conscious environment,return on investment and demonstration of use cases to be showcased in a compelling manner for C-suite to be on board.Operational risk Implementing AI creates operational risk in a

76、 highly regulated,risk adverse industry where operational resilience is highly scrutinised by regulators.Sector use cases Innovation leading to customer valueInnovative AI can bring value to customers both in service offerings and in speed of claim submission.For example,using AI to verify damage ph

77、otos or claim documents can help screen claims for common mistakes before going to human review.Better quality chat bots help customers with questions about their insurance coverage create a more complex and sophisticated offering.Alternatively,innovation in customer-facing products roadside acciden

78、t assistance,for instance could improve the customer experience by aligning customer need with customer service.Geographical footprintThe insurance industry in the UK is geographically dispersed and acts as a local and regional employer.Adoption of back office AI technology will allow insurers to ma

79、intain and bolster their regional presence,embrace remote work,and reap the benefits of a wider talent pool and university cluster.Productivity gainThe adoption of AI can augment revenue,control and support functions such as optimising pricing,better policy administration and claims management,highe

80、r quality data and reporting,and better risk controls.Section 2:The projected impact of AI on the UKs financial and professional services workforce15Financial&Professional Services:The Future of AI&the WorkforceRetail banking:Impact of AIRetail banking is taking a more cautious approach to AI adopti

81、on compared to other sectors and is focusing on getting the fundamentals right in terms of data,infrastructure,and risk management before fully embracing AI-powered decision-making.With a large amount of customer data at risk and uncertainty over regulation,retail banks are integrating AI on a case-

82、by-case basis with many fail-safes in place to prevent failure.Primarily use cases,such as“agent assist”,have proven beneficial to date.AI skills across the retail banking workforce+236%increased demand for AI skills in the last three years There is a significant growth in demand for AI skills acros

83、s retail banking.4.5%The figure below shows the spread of AI skills being seen in job posts in the last 12 months.Please see appendix for methodology.of job posts in the last 12 months include AI skillsGen AI&ChatbotsAI&ML TechniquesRetail bankingAI&ML Core108329100736AI skill groupsJob posts in las

84、t 12-monthsVery fast growth.Either 150%OR any skill with a 0 baseline in 3YA data(i.e.100%growth)Fast growth.Between 50%and 149.9%changeLower growth.Below 49.9%changeAI&ML Governance&OperationsGraph 3:AI skills retail in banking03006009001200150016Financial&Professional Services:The Future of AI&the

85、 WorkforceSection 2:The projected impact of AI on the UKs financial and professional services workforceChallengesLegacy systems Many retail banks are still addressing legacy IT systems and dispersed data across different products and services,making it difficult to access high-quality,integrated dat

86、a needed to effectively deploy AI.Culture Retail banks have focussed on building a culture of strong conduct,therefore being more risk adverse.When it comes to adopting new technologies like AI,by comparison with the“automation-first”mindset of more agile challenger banks,retail banks have been slow

87、er to see the benefits of AI adoption.Controls Many retail banks are starting to trust AI models in approving credit applications without a human in the loop process.Whilst this has benefits,robust controls and fail safes are required to protect the bank against errors in AI judgment.17Financial&Pro

88、fessional Services:The Future of AI&the WorkforceSection 2:The projected impact of AI on the UKs financial and professional services workforceSector use cases Retail banks are taking an incremental approach to AI integration,with the focus on enhancing human decision-making and operational efficacy.

89、Albeit fully autonomous AI is far away,an approach that balances being in favour of AI advancement against intelligence concerning the risks is required.In the long run,when the robustness of the technology satisfies the industrys risk appetite,we will likely see the role of AI increase,for example

90、with fully/semi-automated investment decisions.Client onboarding:Headway has been made in onboarding and Know Your Client(KYC).AI can automate document verification and KYC checks,streamlining the account opening process and ensuring compliance while allowing employees to focus on less admin-intensi

91、ve activities.Fraud detection:Progress with respect to controls has been made.For example,in counter-fraud,AI is now used to detect anomalous transactions and flag potentially fraudulent activities in real-time.Such controls are fast becoming more an expectation than a“nice-to-have”for retail banks.

92、Fintech:Impact of AIAI is a priority on the Fintech corporate agenda.Given the nature of the sector,there is creative licence to experiment and build AI into products.Fintechs have more modern,flexible technology;while some may not have access to large amounts of data,what they have is high-quality

93、and up to date,enabling them more easily to integrate and train AI for financial sector products.AI skills across the fintech workforce+40%increased demand for AI skills in the last three years Fintechs ask for AI skills more than any other FPS sub-sector.5.4%The figure below shows the spread of AI

94、skills being seen in job posts in the last 12 months.Please see appendix for methodology.of job posts in the last 12 months include AI skillsFintechGen AI&ChatbotsAI&ML TechniquesAI&ML Core8018029329Graph 4:AI skills in fintechAI&ML Governance&OperationsAI skill groupsJob posts in last 12-monthsVery

95、 fast growth.Either 150%OR any skill with a 0 baseline in 3YA data(i.e.100%growth)Fast growth.Between 50%and 149.9%changeLower growth.Below 49.9%change03006009001200150018Financial&Professional Services:The Future of AI&the WorkforceSection 2:The projected impact of AI on the UKs financial and profe

96、ssional services workforceChallengesData accessibilityFintechs often still face challenges in accessing and utilising vast volumes of data for AI implementation also known as“Big Data”.Siloed data and reliance on partner infrastructure both restrict the availability of data,while regulatory and risk

97、 concerns limit the use of open-source AI for customer data processing and advice generation.Fintechs need access to larger data sets to enrich insights and power more intuitive AI,and will be developing strategies for improving access to large data pools.First mover advantageSome large players in t

98、he market already boast advanced AI capabilities,with the technology integrated throughout a range of products.Other fintechs must remain competitive by moving fast to foster rapid innovation.19Financial&Professional Services:The Future of AI&the WorkforceSection 2:The projected impact of AI on the

99、UKs financial and professional services workforceSector use cases Product development and financial accessibilityAI has enabled the creation of more innovative and personalised financial products and services,driving growth and competitiveness in the fintech industry.However,as fintechs develop more

100、 AI-reliant products,there will be an increasing focus on building in safeguards against biases to support product effectiveness and accessibility across a variety of social and other characteristics,thereby bolstering diversity and inclusion.Risk and controlAI will enhance various aspects of fintec

101、h control and support operations,including fraud detection,KYC processes,and regulatory compliance.Challenger banks:Impact of AIChallenger banks continue to embrace AI advances to drive their operations and customer experience.These digital-first banks are leveraging AI capabilities such as predicti

102、ve analytics and personalisation to offer quicker,cheaper,and more tailored services to their primarily younger customer base.While large incumbent banks are also investing in AI,challenger banks lean business models with modern data structures allow them to benefit from AI innovations at greater sp

103、eed than traditional banks.AI skills across the challenger banks workforce+196%increased demand for AI skills in the last three years Challenger banks currently hire for fewer AI skills than fintechs,but this gap is quickly closing.4.9%The figure below shows the spread of AI skills being seen in job

104、 posts in the last 12 months.Please see appendix for methodology.Section 2:The projected impact of AI on the UKs financial and professional services workforceof job posts in the last 12 months include AI skillsChallenger banksGen AI&ChatbotsAI&ML TechniquesAI&ML Core4810410413AI skill groupsAI&ML Go

105、vernance&OperationsVery fast growth.Either 150%OR any skill with a 0 baseline in 3YA data(i.e.100%growth)Fast growth.Between 50%and 149.9%changeLower growth.Below 49.9%changeJob posts in last 12-monthsGraph 5:AI skills in challenger banks03006009001200150020Financial&Professional Services:The Future

106、 of AI&the WorkforceChallengesEthical concernsCredit scoring and assessments are ripe for transformation by AI,with applications and approvals for products bypassing human interaction.While this may make for an efficient service,it could simultaneously disadvantage individuals by failing to properly

107、 account for diversity or specific personal circumstances.Big data Large language models(LLMs),models that can achieve general-purpose language understanding and generation rely on large datasets to make tools intuitive and intelligent.In general,challenger banks,being smaller than their rivals,lack

108、 their access to copious data,making it relatively more difficult to pursue new opportunities for broader banking products.RegulationAlthough AI presents new opportunities in enhancing customer-facing processes,caution around data security and regulation,when already highly automated,remains a major

109、 factor leading to human input remaining in the process.Sector use cases Customer services Although fully-automated,AI-generated customer advice is not on the immediate horizon in view of regulatory concerns,structured AI with standard response capabilities is shortening customer service waiting tim

110、es and triaging concerns more efficiently.Challenger banks are seeking to use these tools to ever greater effect while preserving a low risk appetite.Customer onboarding and counter-fraud AI is employed to improve the customer onboarding process by authenticating images used in ID checks.In counter-

111、fraud,AI models aid detection and prevention,with human agents alerted where intervention is required.Productivity AI can summarise customer interactions,allowing agents instead of writing reports to concentrate on resolving issues,i.e.to work on value-add tasks.Section 2:The projected impact of AI

112、on the UKs financial and professional services workforce21Financial&Professional Services:The Future of AI&the WorkforceInvestment banking:Impact of AIInvestment banks have been using machine learning and data science-led algorithms for a long time to support trading.GenAI represents the next step i

113、n that evolution and can lead to a competitive edge.This is,indeed,a welcome development;though investment banks must adopt the technology at a pace which considers regulatory requirements and satisfies shareholder risk appetites.Therefore,many firms are experimenting with new technology in test env

114、ironments(e.g.low complexity and low risk tasks)while at the same time turning to AI to automate processes across trading,risk,and legal.Human oversight is still required for complex products as risks around AI adoption must be managed.AI skills across the Investment banking workforce of job posts i

115、n the last 12 months include AI skills+76%increased demand for AI skills in the last three years AI adoption is starting to grow but AI skills within the workforce are lagging.4.9%The figure below shows the spread of AI skills being seen in job posts in the last 12 months.Please see appendix for met

116、hodology.Section 2:The projected impact of AI on the UKs financial and professional services workforceGen AI&ChatbotsAI&ML TechniquesAI&ML Core4810410413Investment bankingAI skill groupsJob posts in last 12-monthsGraph 6:AI skills in investment bankingVery fast growth.Either 150%OR any skill with a

117、0 baseline in 3YA data(i.e.100%growth)Fast growth.Between 50%and 149.9%changeLower growth.Below 49.9%changeAI&ML Governance&Operations03006009001200150022Financial&Professional Services:The Future of AI&the WorkforceChallengesRegulatory environmentThere is a cautious approach to AI adoption in large

118、r institutions due to the regulatory environment.Upcoming regulations around fines and recalls have made banks more reserved in their AI ambitions.Explainable AIAs investment banks increasingly automate decision-making to operate at a faster pace,there is a growing requirement to explain the rationa

119、le behind AI-driven investment decisions to stakeholders and regulators.Transparent and interpretable AI models in addition to technological guardrails are crucial.Sector use cases Automation and efficiencyAI and machine learning technologies are enabling investment banks to automate a wide range of

120、 routine tasks and processes,from trading and risk management to legal documentation to End-User-Computing.This permits investment banks to achieve higher levels of efficiency while freeing up human experts to focus on more complex,value-add activities such as building client relationship and innova

121、ting processes.Enhanced insights and decision-makingAI-powered tools are providing investment bankers with advanced analytics and predictive capabilities,enabling them to extract valuable insights from large datasets and to make more Risk AppetiteAlgorithmic trading (rules based computational formul

122、as for executing trades)and taking automated positions are already within the investment strategy of investment banks but human validation and execution is still present.These institutions are aware of the effect on the broader market if AI automated trading is deployed incorrectly presenting a barr

123、ier to full scale automation of trading.”23Financial&Professional Services:The Future of AI&the Workforceinformed,data-driven decisions in areas including sentiment analysis,news-driven trading strategies,and personalised customer offerings.AI-aided investmentsIn the future,when investment bankers h

124、ave honed the ability to keep abreast of iterative improvements to AI,there could be scope for the technology to aid clients with investment decisions based on AI generated prompts,lowering the cost of investment advice and thereby rendering it more accessible.Legal services:Impact of AIThe adoption

125、 of AI is fast becoming a field of expertise in its own right for lawyers,as advice sought by clients on the legalities of implementation rises markedly.Yet use of the technology within the profession has been held back by concerns relating to data and explainability.As such,adoption remains confine

126、d to relatively small,specific areas of the business while being kept in check by human-in-the-loop oversight.That being said,seeing as law is a language in/language out business,generative AI tools like ChatGPT can significantly enhance ways of working in the sector.Use cases around supporting volu

127、me-based tasks and admin-heavy processes such as contract writing are exciting for the legal practice.AI skills across the Legal services workforce of job posts in the last 12 months include AI skills+345%increased demand for AI skills in the last three years Legal services currently ask for the few

128、est AI skills,but this is changing.1.2%The figure below shows the spread of AI skills being seen in job posts in the last 12 months.Please see appendix for methodology.Section 2:The projected impact of AI on the UKs financial and professional services workforceGen AI&ChatbotsAI&ML TechniquesAI&ML Co

129、re30421363Legal servicesVery fast growth.Either 150%OR any skill with a 0 baseline in 3YA data(i.e.100%growth)Fast growth.Between 50%and 149.9%changeLower growth.Below 49.9%changeAI skill groupsJob posts in last 12-monthsAI&ML Governance&OperationsGraph 7:AI skills in legal services03006009001200150

130、024Financial&Professional Services:The Future of AI&the WorkforceSector use cases Self-service legal products and toolsAI tools that are capable of interpreting the law to answer legal questions or completing such tasks as drafting contracts will help speed up and make less complex legal work more a

131、ccessible.Tools like this will require closer monitoring from a risk perspective as well the recruitment of tech talent to be maintained.Evolving business modelsThe traditional time-and-materials billing model in legal services will likely give way to more ChallengesExplainabilityThe essence of what

132、 a lawyer does is to interpret the law in the light of specific circumstances,and to cogently explain the thought process that led them there.It follows that AI tools in the profession must be capable of detailing how they arrive at their conclusions for the technology to pass muster.Lawyers of the

133、futureA shift in skillsets will be mandated as the legal practice moves into the future,with technology and AI literacy soon becoming foundational.For all that automation will do,with the traditional analytical tasks of a lawyer,professionals will still be required to interpret the outputs of AI,rev

134、iew the thinking,and present to their clients.This will likely call for more interpersonal and strategic thinking competencies among junior staff members.Grassroots programmes(traineeships,graduate schemes,and apprenticeships)will all need updating to reflect the centrality of AI skills among traine

135、es moving forward.Data PrivacyThere is a risk of sensitive information being leaked and therefore lawyers are hesitant or even blocking the use of GenAI currently.outcome-based,value-driven pricing as AI drives efficiency.Law firms will need to redesign their cost structures and staffing models to d

136、o more with fewer lawyers as AI automates an increasing share of legal work.In instances of time and ma-terial cost structures,AI will carve out potential for law firms to seize new business opportunities where,historically,charge-out rates per lawyer had been unprofitable.Section 2:The projected im

137、pact of AI on the UKs financial and professional services workforce25Financial&Professional Services:The Future of AI&the WorkforceRevenue functions:Impact of AIWhilst AI impact projections for revenue functions are below those for control and support by 10%and 20%respectively(driven by a greater pr

138、oportion of revenue-generating work requiring personal interaction with customers),the opportunity to free up time spent on operational and administrative activity is substantial.The greatest potential gains are in heavily transactional areas such as Sales&Trading,as well as areas that involve summa

139、rising information,conducting research,and preparing content,such as Strategy&Insights and Project Management.Workers in revenue functions may need to upskill in attention to detail and quality assurance to ensure that generated content is fit-for-purpose.Workers in Tech Product in particular will r

140、ely on prompt engineering to create more sophisticated segments of code,and those engaged in trading activities should upskill in core AI concepts so they can critically understand trading recommendations and risk outputs.It is also important to note that this analysis focuses on augmenting the sect

141、or as it is today.The opportunity to leverage AI to access new markets and offer new hyper-personalised products and services may offer even greater potential gains,but are still too immature to be quantified.Upskilling in continuous improvement,commercial acumen,and product will all contribute to t

142、he sector capitalising on new opportunities for AI and Gen AI powered products&servicesTask groupsCommentary&use casesSales&trading activities will be most impacted across bulk program trading activities where code generation will boost quant/algorithmic developers,although traditional sales&trading

143、 activity can be accelerated via enhanced portfolio tracking,market analysis and recommendations.Sales&TradingTech ProductTech product could be accelerated in code generation,survey design,analysis of user feedback,requirements gathering,user story generation,and backlog management.Project Managemen

144、tAI-enhanced project management will augment tracking of project performance and outcomes,communicating project activities,project planning,and identifying and mitigating risks.Strategy&insights boosts may come via acceleration of market research petitor analysis,simulation of potential outcomes,str

145、ategy ideation,processing of customer data,creating project plans,and automating tracking of KPIs.Strategy&Insights14%14%14%10%39%38%37%43%Section 2:The projected impact of AI on the UKs financial and professional services workforceProductivity boost year 1Productivity boost years 2-5(5 year total:5

146、3%)(5 year total:53%)(5 year total:52%)(5 year total:51%)26Financial&Professional Services:The Future of AI&the WorkforceAccount ManagementFinancialProductsSalesLegal AdvisoryFinancialAdvisory Relationship ManagementAccount management may be boosted as drafting communications,scheduling of interacti

147、ons and auto-responses accelerates the account manager.Financial products are likely to be boosted across training,marketing,development and compliance.Flexible pricing strategies and competitor analysis will also contribute.AI can enhance sales activity by supporting minor tasks like preparing cont

148、act lists,helping prepare agreement drafts for review and nudging cross sales and up sales.Legal advisory AI can assist with research activities,provide some suggestions for advice,and help make amendments and suggestions for legal documents.However,a legal professional must approve all final docume

149、nts.Financial advisory may see AI perform calculations and make better data-driven forecasts.Potential to provide more personalised recommendations and perform better quality holistic risk assessments.Relationship management is(unsurprisingly)the most insulated from AI,although it may support with a

150、ddressing client queries,generating faster comms&client specific content,and scheduling/nudging for follow ups.Customer SupportConversational AI designed to triage and resolve customer support requests,summarise content,work in multiple languages,and offer personalised support are likely to transfor

151、m customer experience.14%9%12%10%13%12%7%25%26%20%33%37%32%32%Section 2:The projected impact of AI on the UKs financial and professional services workforceProductivity boost year 1Productivity boost years 2-5Investment ManagementInvestment management may see auto-compiled investment&market performan

152、ce reporting,dynamic asset recommendations,and enhanced checks to ensure trades are within policy parameters.14%34%(5 year total:48%)(5 year total:47%)(5 year total:46%)(5 year total:44%)(5 year total:42%)(5 year total:38%)(5 year total:38%)(5 year total:27%)27Financial&Professional Services:The Fut

153、ure of AI&the WorkforceControl functions:Impact of AI Appropriate management of risk sits at the centre of FPS firms ability to operate,ensuring customers and the wider public has sufficient trust in the sector for it to operate effectively.With the contemporary geopolitical,environmental,social and

154、 technological disruptors in the global economy,we are seeing the level and nature of enterprise risk grow in magnitude and complexity.FS firms can leverage AI to help meet this challenge,enhancing control functions to be more comprehensive,accurate and effective.AI also presents its own risks as we

155、ll as a new and evolving set of regulations to adhere to.We may see significant growth in some control areas as AI governance capacity expands to meet demand.More broadly,we are also likely to see work increasingly move into an approver/validator space as operational tasks are automated but humans a

156、re accountable for a broader range of generative AI outputs that need checking and approving.Skills in transformation and business process design will be important for the generative-AI transformational phase within control functions.Control function professionals will also need to upskill in core A

157、I skills to help them spot hallucinations(a symptom of the inherent creativity of large language models that sometimes causes them to give false information),and prompt engineering to help extract,summarise and compare policies,regulatory and legal documents in different contexts-all with substantia

158、l attention to detail.Control functions will increasingly need to upskill in data architecture,quality and concepts so that appropriate judgements can be made regarding the validity of automated data-driven outputs and recommendations and give appropriate recommendations about storing and processing

159、 sensitive customer data.Hallucinations are a particular area of attention for senior leaders in control functions implementing generative AI solutions.An anti-money laundering check that falsely identifies an issue or a generative AI process incorrectly impacting a credit score could carry substant

160、ial reputational risk for firms.Leaders should consider multiple lines of defence to mitigate against this both technological and human.The importance of appropriate controls in the era of generative AI cannot be understated.FS firms have full access to real-time IT security data from across their e

161、state,making it a perfect candidate for AI-enhancement.The most valuable Information Security use cases involve identifying suspicious network patterns,automation of testing,extracting and summarising system data,regular reporting and even suggesting mitigation strategies.Commentary&use cases Inform

162、ationSecurity11%55%Section 2:The projected impact of AI on the UKs financial and professional services workforceProductivity boost year 1Productivity boost years 2-5Task groups(5 year total:66%)28Financial&Professional Services:The Future of AI&the WorkforceAuditLegal&Compliance 12%10%36%30%Establis

163、hed and repeatable methodologies that tend to underpin audit operations offer good candidates for AI.AI can accelerate completion of audit checklists,ensuring that certain requirements have been met in company documents&external disclosures.Legal&compliance operations tend to involve the implementat

164、ion of internal and external regulatory/compliance frameworks and the interpretation of legal documents all great candidates for AI.For legal and compliance tasks,AI can compare digital information against any company legal and compliance framework and escalate any issues found for human review.Time

165、 consuming tasks such as legal consultation,regulatory liaison and litigation management will remain significantly human-led,driving overall lower impact scores for legal&compliance.See page 9 for details of parameters of productivity boost figures.Section 2:The projected impact of AI on the UKs fin

166、ancial and professional services workforceProductivity boost year 1Productivity boost years 2-5Risk Management15%38%Risk management involves triangulation of data from multiple sources and comparison against organisational thresholds for risk,presenting good opportunities for AI.Value-add use cases

167、involve organising risk models and processing risk-related data,capturing risk data efficiently and triggering alerts for immediate attention.AI can also automate parts of data extraction and preliminary assessments,while aggregating and analysing strategic information to enhance proactive risk repo

168、rting.(5 year total:53%)(5 year total:48%)(5 year total:40%)29Financial&Professional Services:The Future of AI&the WorkforceSupport functions:Impact of AI Support functions in FPS stand to undergo an estimated 60%productivity boost over the next 5 years,driven by the automation of bulk operational p

169、rocesses,user support via conversational dialogues,and code generation in IT.Workers will have more time freed up to focus on strategic&business-partnering activities,as well as expanding the capacity and reach of their operations.As Support Function services transform to embrace generative AI over

170、the coming years,upskilling in continuous improvement and growth mindset will help workers thrive amidst a period of change.Knowledge of process engineering will give functional experts the opportunity to contribute to better business processes,while data architecture and prompt engineering will be

171、the biggest upskilling area for IT as greater sophistication is built around enterprise data management,and the code creation process is increasingly automated via well-written prompts.HR and Finance workers in particular might upskill in data security and privacy principles due to the sensitivity o

172、f employee and customer data in their domains,while Marketing and Communications can upskill in the content creation capabilities of generative AI across text,audio,image and video to revolutionise the productivity and reach of their work.Within IT,for Data Management,AI can catalogue data,automate

173、data quality checking and complete integrity checks,and consolidate&reconcile data across multiple sources.The most impact in IT is likely to come from code generation and translation capabilities,which may also substantially decrease barriers to entry for new IT professionals.Commentary&use cases I

174、T16%55%Section 2:The projected impact of AI on the UKs financial and professional services workforceProductivity boost year 1Productivity boost years 2-5Task groups(5 year total:71%)30Financial&Professional Services:The Future of AI&the WorkforceFor marketing campaigns and activities,AI can lend its

175、elf to drafting of content in multiple languages.AI digital marketing tools can enhance visual,video and audio-based content and also help with distribution using automation strategies across search and social media.AI can be used in the ideation phase all the way through to scheduling the release o

176、f content.Communications channels can become more inclusive with AI technologies and large language models,including translating between languages and drafting internal and external communications in the tone of the reader.HR can expand its reach into the businesses and focus more on business partne

177、ring by using AI to free up time in the recruitment process across candidate sourcing and selection,dissemination of better information during employee onboarding,dynamic talent mobility platforms,proactive distribution of strategic learning,and more dynamic ways to listen to employees and gather se

178、ntiment.Marketing&CommunicationsHR16%11%38%35%Operations15%45%Operations rely on high level of data extraction(e.g.from customer documents),data entry(e.g.into company systems),validation checks(paring data with frameworks and thresholds),and customer updates.All of these can be significantly accele

179、rated using AI.Furthermore,language translation will give organisations more options around where to locate operational centres.Finance15%43%Finance operations may soon involve AI-accelerated tasks such as accounts generation,P/L reporting,form filling,and report creation-including drafting of regul

180、atory disclosures.AI can be used to accelerate processes,minimise errors.AI can aid budget management and financial forecasting through data processing and report generation.Section 2:The projected impact of AI on the UKs financial and professional services workforceProductivity boost year 1Producti

181、vity boost years 2-5(5 year total:58%)(5 year total:54%)(5 year total:46%)(5 year total:60%)31Financial&Professional Services:The Future of AI&the WorkforceSection 3What employers should consider to maximise the benefits of using AI in their organisationsA framework for maximising the benefits of AI

182、This methodology shows how AI integration should be navigated with actions which will support both a culture shift and training to aid employees in embracing this technology.By following a systematic or structured/strategic approach,organisations can navigate the complexities of AI adoption,address

183、potential challenges,and maximize the benefits derived from AI applications.01Communicate the visionClearly communicate the companys vision for adopting AI and how it aligns with the organizations goals.Explain the potential benefits and how AI will enhance employees work.Make it clear02Establish et

184、hical guidelines and risk managementDevelop clear ethical guidelines for using AI in the workplace.Ensure that employ-ees understand and adhere to these guidelines to maintain trust and transparency.03Launch the AI toolLaunch the AI tool with the message coming from the top of the company and with l

185、eadership role-modelling the change.Make it known04Provide training and supportOffer comprehensive training programs to help employees understand the AI tool,its capabilities,and how to use it effectively.Provide ongoing support to address any questions or concerns.05Encourage collaboration and expe

186、rimentationFoster a culture of collaboration between employees and the AI tool.Encourage employees to provide feedback,suggestions,and ideas for improving the tools functionality.Allow employees to experiment with the AI tool and explore its potential applications in their daily tasks.This encourage

187、s creativity and innovation.Make it real06Showcase success storiesShare real-life examples of how AI has successfully improved work processes,enhanced productivity,and achieved positive outcomes within the company.07Address concernsAcknowledge and address employees concerns about job displacement,pr

188、ivacy,and ethical implications of AI.Provide transparent information and assure employees that AI is meant to augment their work,not replace them.Make it happen08Measure and communicate resultsRegularly assess the impact of AI on productivity,efficiency,and overall business performance.Share these r

189、esults with employees to demonstrate the positive out-comes of AI adoption.09Foster a learning cultureEncourage employees to continuously learn and adapt to the changing landscape of AI.Provide opportunities for professional development and upskilling related to AI.Make it stick10Monitor and adjustC

190、ontinuously monitor the implementation of AI and gather feedback from employ-ees.Be open to making adjustments based on the experiences and suggestions of the workforce.Section 3:What employers should consider to maximise the benefits of using AI in their organisations33Financial&Professional Servic

191、es:The Future of AI&the Workforce1.Creating an AI strategy AI is already being leveraged by firms for a multitude of purposes such as improving customer service,process automation,and augmenting personal productivity.With this in mind,firms need to have a comprehensive AI strategy which takes into c

192、onsideration workforce skill requirement,regulation,risk appetite,competitive edge,customer expectations,and profitability,whilst considering and curating organisational culture,habits,and work patterns.Alongside this,firms need to consider the procurement of AI in a“buy and/or build”capacity do we

193、focus on developing internally,sourcing externally or an amalgamation of both.Consider:AI adoption should focus on priority use cases and initiatives that deliver rapid and impactful return on investment(ROI)across the enterprise including both internal and external value.Identifying areas where AI

194、can provide a competitive advantage and meet evolving customer expectations as well as the potential benefits and impact on workforce and process productivity.The strategy must include clear alignment on regulation and risk appetite with a robust evaluation process to monitor the effectiveness of th

195、e AI initiatives.PurposeProtect People PacePrepareEmployer considerationsHaving the right framework and risk metrics in place to protect the firm and its data.Firms should put in place a central governance body responsible for overseeing GenAI innovation and guiding employees through the AI transfor

196、mation,thereby driving the development of new business and service delivery models powered by AI.Identifying the workforce implications of AI adoption on role augmentation,and requirements for upskilling,reskilling and redeployment in growth areas for the business.Ensure your foundations are in plac

197、e including establishing a robust data infrastructure and governance framework for ensuring data quality,security,and accessibility.In addition,plan and implement effective change management strategies with a clear communication strategy,driving employee buy-in and minimising a resistance to AI adop

198、tion.Understand your pace of adoption and the factors which will determine this.This includes the rate of upskilling,regulatory expectations,technology development and your business risk appetite.Whilst AI isnt something that can be opted out,moving too fast could create unnecessary risk.Section 3:W

199、hat employers should consider to maximise the benefits of using AI in their organisationsWhile the implementation and adoption of new technologies is nothing new for FPS firms,as AI continues to grow in use,there are a number of considerations employers need to prioritise and for which they must put

200、 clear strategies in place.34Financial&Professional Services:The Future of AI&the Workforce2.Three paths to deploying AIThe cross-functional nature of AI is creating some ambiguity within organisations with regard to who is responsible for AI and what the roadmap to AI transformation should be.Using

201、 these key paths will enable the sector to understand the stages of evolution and allow for easier adoption.The tables below show the productivity increase,risk level,platform complexity,and change/time required for each path of AI.1.General augmentation General personal augmentation(GPA)is a suite

202、of productivity tools that automate common tasks for knowledge workers.These tools are embedded in existing software and websites to help with document creation,email drafting,data summarization,and content generation.They are designed for ad-hoc use by individual workers rather than for enterprise

203、processes.GPA solutions are readily available“off the shelf”and easy to implement.However,their flexibility also increases their risk profile.Users have a high degree of freedom to use them but they also require significant behaviour change.While GPA solutions are the easiest to implement,they have

204、lower productivity returns and require significant user adaptation.Productivity increaseRisk LevelPlatform complexityChange/time requiredLOWMEDIUMLOWMEDIUMProductivity increaseRisk levelPlatform complexityChange/time requiredMEDIUMLOWMEDIUMLOWProductivity increaseRisk levelPlatform complexityChange/

205、time requiredHIGHHIGHHIGHHIGHSection 3:What employers should consider to maximise the benefits of using AI in their organisations2.Enhanced process automation Enhanced process automation(EPA)leverages moderately complex AI applications to streamline and transform existing processes.This includes aut

206、omating tasks like tier 1 support and disseminating policy information.In the future,EPA applications may handle more personal information and provide higher-value outputs,such as automated trading recommendations or suspicious pattern detection.While EPA offers significant productivity gains and co

207、ntrolled risk through process controls,it requires a more sophisticated technical infrastructure,governance,and AI strategy,leading to higher platform complexity.3.Enterprise augmentationEnterprise augmentation marks a shift from legacy human-centric processes to AI-enabled organizations with reimag

208、ined processes.Humans remain critical,directing,controlling,and curating AI outputs while applying expertise for strategic context.Reaching this level of maturity requires years of development due to the complexity of supporting infrastructure and the significant change involved.However,it is necess

209、ary to fully realize the potential gains of generative AI across the value chain.35Financial&Professional Services:The Future of AI&the Workforce3.Workforce planning,reskilling,and mobilityThere is no doubt that AI and generative AI in particular will continue to enhance the way work is done across

210、FPS firms.It will fuel productivity gains and free up time for colleagues to focus on higher-value activities which drive growth.As FPS firms recognise the benefits of AI and see this translated into greater productivity,there will generally be four options to consider:1.Re-allocation of saved time

211、to high-value tasks within the scope of the role 2.Redeployment of talent into areas of the business with high need and growth 3.Maintaining the same structure,but with less stress/pressure due to time recouped through increased productivity 4.Redundancy and short-term cost savings for roles wherein

212、 the majority of tasks can be replaced by AI N aturally,the fourth option will appear an attractive proposition to cost-sensitive firms.The reality,however,is that while AI may be able to augment and replace many tasks,it will also create new jobs for employees requiring new skills.The Financial Ser

213、vices Skills Commission(FSSC)found that financial services firms could save up to 49,100 per employee when they reskill an employee,compared to hiring someone new with the relevant skills12.Organisations channelling AI primarily through a cost-cutting prism may prove short-sighted and reactionary in

214、 doing so.Instead,proactively upskilling and reskilling colleagues in higher-value activities and growth areas of the business will lead to real and long-term cost savings and financial gains.In line with expected workforce changes,organisations must build a data-led view of the skills they currentl

215、y have,the skills they will need in future(with an eye to the implications of AI),and the gaps.This will enable targeted learning and development and sourcing of talent to offset critical shortages.The FSSC Future Skills Framework provides a good place to start for firms looking for common future sk

216、ills which will be required within the financial services sector.Many of the AI and AI-adjacent skills of the future are still relatively new and not yet prevalent in the labour market.Therefore,employers should consider bridging the AI knowledge gap via training and development rather than by looki

217、ng to recruit AI talent in the scarce and highly competitive marketplace.The Skills and Talent section of the UK Digital Strategy supports this with a focus on AI upskilling,addressing general digital skills gaps through initiatives such as the Digital Skills Council and Skills Bootcamps,as well as

218、encouraging employer-led upskilling13.Greater emphasis on specialised training programmes for advanced AI skills would further enhance the development of a workforce capable of driving AI innovation,as would introducing substantive incentives for employers to invest specifically in AI training.AI dr

219、iven workforce planning should be dependent on the pace of adoption and the growth of the company.Most companies that have successfully come through AI implementations and,as a result,freed up capacity,are focusing additional time on improving their customer offering,whether that be through shorter

220、waiting times for call centres or by working on new product ideas.Where AI is adopted more gradually,companies may opt for increasing volume per employee or for simply not plugging the gaps left by natural attrition,as opposed to realising short-term benefits by way of redeployment or redundancy pro

221、grammes.Leaders wishing to prime themselves for growth in the medium term should favour investment in reskilling over reflexive redundancies.”Mel Newton,KPMG People and Change Partner“Section 3:What employers should consider to maximise the benefits of using AI in their organisations36Financial&Prof

222、essional Services:The Future of AI&the Workforce4.Regulation and operational resilience As AI is data driven with a degree of autonomy,it has the potential to pose a high risk if not managed correctly.It is important firms have the correct infrastructure and governance set up before moving forward w

223、ith AI applications.Regulation Regulation and legislation for AI is rapidly developing worldwide.The EU AI Act represents a significant step in establishing stringent regulations.In the UK,efforts are focused on promoting a pro-innovation response by adopting a risk-based approach and offering guida

224、nce based on responsible AI principles,while leveraging existing regulations.There is recognition of the need for a review to address any identified regulatory gaps and to enhance guidance,particularly in relation to specific legislative requirements such as explainability in the context of GenAI.Op

225、erational resilienceRobust strategies should be enacted to ensure operational resilience in cases of AI systems failure.This includes drawing up contingency plans,installing backstops,and guaranteeing the ability to quickly detect and respond to AI errors.Data governance and strategyEstablishing cle

226、ar governance policies and strategies for safeguarding the quality,integrity,and security of data used in AI systems is critical.However,in addition,employers must also consider the growing importance of broader AI governance.This encompasses not only these foundational aspects of data management,su

227、ch as defining data ownership,setting data quality standards,and implementing processes for data lineage and cataloguing,but also the need for AI Boards or Councils.These bodies bring together diverse perspectives to collaboratively oversee and challenge the ethical and safety implications of AI use

228、,ensuring a comprehensive approach to governance.Leanne Allen KPMG Partner,Data,Data Science and AI Lead“In a world increasingly powered by AI,responsible development and deployment are not only ethical imperatives,its the foundation for building trust and securing long-term business success.”Sectio

229、n 3:What employers should consider to maximise the benefits of using AI in their organisationsFinancial&Professional Services:The Future of AI&the Workforce375.Responsible AIHarnessing this potential of AI requires a commitment to responsible,ethical,and trustworthy AI development and deployment.Com

230、pliance with all applicable regulations should be rigorously adhered to,but beyond compliance,its about embedding ethical principles like fairness,transparency,and accountability into the very fabric of AI initiatives.This is crucial as public trust will drive AIs adoption success.A robust,responsib

231、le AI framework serves as a guidepost,helping mitigate risks across key principles:fairness,privacy,transparency,sustainability,explainability,data integrity,accountability,reliability,security,and safety.Prioritising these principles can ensure that AI becomes a force for good,driving innovation,wh

232、ilst protecting brand reputation,enhancing customer loyalty and building trust.Tech guardrails and architecture LLMs are powerful but complex tools.Their very nature makes them susceptible to the pitfalls of poor prompting.Even subtle variations in questions or data inputs can lead to wildly diverge

233、nt and inaccurate outputs which poses significant risk,as evidenced by lawsuits such as one filed against Air Canada when its support chatbot gave inaccurate information to a grieving customer.Implementing robust technical guardrails and architectures can effectively prevent the dissemination of inc

234、orrect or misleading responses.These safeguards should encompass elements as evidenced by recent lawsuits such as:Education on Prompt Engineering:Developing guidelines for crafting clear,unambiguous prompts that can minimise ambiguity and bias.Input Validation:Introducing privacy and security mechan

235、isms to validate and sanitise user inputs.It is vital that decision-making filters out potentially harmful,personal,sensitive or misleading information.Whats important is that we are solving problems for business and focusing on where AI can improve the experience for our employees and customers”Pen

236、ny Jones Zurich,Responsible AI Lead“Section 3:What employers should consider to maximise the benefits of using AI in their organisations38Financial&Professional Services:The Future of AI&the Workforce Output Monitoring:Continuously monitoring LLM outputs for inconsistencies,factual errors,and potent

237、ial biases.Human-in-the-Loop:Integrating human oversight into any critical decision-making processes to drive responsible and ethical AI responses.AI hallucinationsEmployers should be aware of the potential for AI systems to produce hallucinations(plausible sounding yet factually inaccurate results)

238、.There are a few mitigating measures that can minimise the risk including:Incorporating human-in-the-loop processes to ensure that AI outputs are reviewed and validated before being implemented.Regular monitoring and evaluation of AI models for signs of hallucination by comparing model outputs with

239、ground truth data.Fine tuning GenAI models on trusted high-quality data to provide pointed relevant responses for your business needs.Right to explainability Firms must be clear how AI tools have come to the conclusions that feed into the choices made.It is the job of employers to ensure that AI sys

240、tems are both transparent and explainable so as to be compliant with legislation such as GDPR.Section 3:What employers should consider to maximise the benefits of using AI in their organisations39Financial&Professional Services:The Future of AI&the Workforce6.Equality,diversity and Inclusion(EDI)AI

241、has the potential to both improve and hinder EDI in FPS services.Customer On the customer side,AI can enable more inclusive financial products and unlock new opportunities for data-driven financial tools to empower a broader range of people.Workforce On the workforce side,younger workers may be bett

242、er able to exploit AI opportunities compared with an aged workforce which struggles to adapt.Low paid and unskilled workforces can benefit from AI enhancing their productivity more quickly,though,could be left behind if training is not available.Accessibility AI can be leveraged in various ways to e

243、nhance accessibility and create more inclusive digital experiences for neurodiverse individuals or those with disabilities.AI-driven hiring tools can mitigate biases,opening up opportunities for diverse candidates.On the flipside,AI risks perpetuating the same prejudices if fed a data pool or deploy

244、ed in such a way that does not embody and enshrine inclusive practices.Socio-economic Diversity Policies to enhance labour mobility(the ability to move from one position to another),address skills mismatches,and provide robust social safety nets will be critical to ensuring AI improves socio-economi

245、c diversity rather than entrenching existing divides.For financial services to promote EDI,AI strategy teams must be diverse and include inclusive mindset to ensure they partner in the evolution with all diverse groupings.Organisations should nominate AI champions who bring out the training for AI b

246、ut also bring together thought patterns,ideas and concerns of their diverse colleague cohort.7.Culture&capabilityWhile the prevalence of AI outside the workplace is high,and people are becoming increasingly accustomed to interacting with and using it,employers should still consider the cultural impl

247、ications of introducing AI into their firms.Most employees tend to fall into one of three categories in respect of their attitudes to AI:1.Unaware of how to use AI;2.Aware but averse to the change(change fatigue resulting from constant innovation in ways of working);3.AI advocates who are both aware

248、 and unafraid.Organisations should stress that AI can better the human experience at work.AI-powered tools like intelligent assistants and chatbots can enhance collaboration,knowledge sharing,and accessibility within the sector,promoting a more inclusive culture.As AI gives a boost to employee produ

249、ctivity,self-efficacy,and job satisfaction by automating mundane tasks,the need arises for deeper strategic work which often requires collaboration.The Organisation for Economic Co-operation and Development(OECD)AI surveys found that more than half(63%)of AI users in finance and manufacturing report

250、ed that AI had improved their job satisfaction14.Placing a greater emphasis on specialised training programmes for advanced AI skills would further enhance the development of a workforce capable of driving AI innovation.Businesses will benefit from this new culture.For example,a career in the financ

251、ial services sector may become a more attractive choice,opening the talent pipeline to a new type of candidate who would previously have disregarded the sector.New entrants contribute cultural capital to the industry,which in turn becomes composed of a data-driven,tech-literate,and strategic yet inc

252、lusive workforce.Section 3:What employers should consider to maximise the benefits of using AI in their organisations40Financial&Professional Services:The Future of AI&the WorkforceAppendixFinancial&Professional Services:The Future of AI&the WorkforceMethodologyTo produce this report,we conducted bo

253、th primary and secondary research to explore the opportunity areas for AI in FPS sectors and a review of the skills and capabilities required for the future.We used public and KPMG internal literature,our network of industry leading subject matter experts and have access to task automation inference

254、 AI tooling to qualify and quantify the impact of AI on the FPS sector.Literature review There is more public interest in AI,and therefore related literature,than ever before.We therefore conducted a literature review of topical thought leadership on AI and its application within the sectors focused

255、 on in this report.This included reports published by KPMG and external trusted sources.Task automation inference AI tool Using a KPMG task automation AI tool,we analysed over 400 tasks from across the FPS sector to assess the extent to which they could be augmented across 7 AI capabilities:content

256、extraction,content creation,summarising information,data processing,conversational dialogues,calendar management&scheduling,code generation,and translation.The total opportunity was summarised by sector and functional area.Subject matter expert (SME)interviews and validationKPMG,Financial Services a

257、nd Artificial Intelligence SMEs were interviewed for their insight into the application of AI in their respectful field of expertise.Our literature review and thought leadership has been validated by the SMEs and all information has been independently reviewed and verified by subject matter experts

258、before publication.42Financial&Professional Services:The Future of AI&the WorkforceProductivity gain assumptionsWe have projected expected productivity gains for the FPS sector as a whole,as well as for revenue generating functions,control functions and support functions using the following set of a

259、ssumptions:1-Year Projection(e.g.by 2025)productivity gains are based on firms implementing and adopting 1“general augmentation”AI product firmwide(phase 1 as defined on the graph below)5-Year Projection(e.g.by 2030)productivity gains are based on firms moving through phases 1,2 and 3.This analyses

260、as-is business operations and does not include projections for new products,services&markets as the technology is too new to make reliable quantitative estimates in this area.Phase 1:General Augmentation(2024/25)Workers can achieve a 10-20%productivity boost across a broad range of day-to-day tasks

261、e.g.drafting documents and summarising spreadsheets.Phase 2:Enhanced Process Automation(2026/27)Increased automation of steps within existing volume enterprise processes e.g.customer onboarding,helpdesk triage/resolution,application processing&data entry.Phase 3:Enterprise aumentation(2029/30)In the

262、 longer term,organisations will transform end-to-end processes from human-centric to AI-centric.Mature AI infrastructure and governance could also support profitable new products,services and markets.Please see Three Paths to Deploying AI within Section 2 for more details on the stages outlined with

263、in the below graph.43Financial&Professional Services:The Future of AI&the WorkforceLabour market analysis methodology Data from job adverts has been used to understand the changing role and skill requirements across the UK FPS sector.The analysis was conducted by Simply Get Results(),a leading skill

264、s data analytics provider,using data from Lightcast(2024).The analysis covers 188k job adverts from nearly 7,000 employers in the UK.The primary focus was to analyse how demand for skills and roles has changed in relation to a range of AI technologies,methods and capabilities.Job adverts were compar

265、ed between two 12-month periods:April 2023 to March 2024 versus April 2020 to March 2021.This three-year span captures meaningful changes while maintaining relevance.All jobs tracked in the labour market data set were grouped into one of three business areas based on the primary focus of their role:

266、Revenue-generating,Control functions and Support.From a skills perspective,a group of 168 technical Artificial Intelligence and Generative AI skills were assessed.Examples include AI Operations,Machine Learning and Large Language Models.This list was supplemented with 17 supporting or enabling skill

267、s which commonly co-occur with the AI skills,such as Analytical Thinking,Decision Making and Quick Learning.The percentage change calculations between the two time periods in this report are prepared based on total volume of demand in job postings in each period.See image for an example.This method

268、was chosen due to the low volume of Generative AI skills in the labour market to date.The number of unique job postings containing one or more relevant skill were also calculated.One job post may contain multiple references to relevant skills.AI Skill GroupAssociated SkillsGen AI&ChatboxChatbotChatG

269、PTCortanaGenerative Adversarial NetworksGenerative Artificial IntelligenceIntelligent AgentIntelligent Virtual AssistantLanguage ModelLarge Language ModellingMulti-Agent SystemsOpenAI GymPrompt EngineeringTransformer(Machine Learning Model)Associated skills for AI skill groups listed in Section 1 ar

270、e outlined here:44Financial&Professional Services:The Future of AI&the WorkforceAI Skill GroupAssociated SkillsAI&ML Techniques(Top 10 skills from 137 identified)Scikit-Learn(Python Package)Deep LearningTensorFlowPyTorch(Machine Learning Library)Random Forest AlgorithmMachine Learning AlgorithmsArti

271、ficial Neural NetworksFeature EngineeringGradient BoostingReinforcement LearningAI&ML Governance&OperationsAIOps(Artificial Intelligence For IT Operations)Applications Of Artificial IntelligenceExplainable AI(XAI)Machine Learning Model Monitoring And Evalua-tionMLOps(Machine Learning Operations)Mode

272、lOpsEthical AIAI&ML CoreArtificial IntelligenceMachine Learning45Financial&Professional Services:The Future of AI&the WorkforceStep 1:Deconstruct work in different FPS business contexts(example below for asset management)Task GroupTaskTask Definition%TimeRationaleLegal&complianceControls TestingTest

273、ing the effectiveness of existing compliance controls within the company10%AI can scan systems for evidence of adherence or contraventions to active policiesPolicy EnforcementMaintain and enforce the firms internal policies regarding financial activities23%AI can assist in policy enforcement via rea

274、l-time monitoring and discrepancy reportingDrafting Legal DocumentsCreating necessary legal documents for the companys operations20%AI could assist in drafting,reviewing,and editing legal documentsRegulatory Thematic ReviewsPerforming in-depth reviews of specific areas of regulatory interest10%AI co

275、uld perform analytics and generate insights for thematic reviewsRegulatory ComplianceEnsuring adherence to financial regulations and laws29%AI can aid with summarisation or regulatory documents and and comparison of company data against regulatory thresholdsLicensing and RegistrationEnsuring the fir

276、m complies with necessary licensing and registration requirements10%AI can create&manage a licensing inventory,monitor licenses and registrations tracking and access changes in licensing requirementsTask analysis methodology KPMG have an established approach to calculating the impact of generative A

277、I on the workforce in different contexts.This involve deconstructing work into its constituent tasks and then estimating the parts of each task that can be automated versus those that will remain human.Below is a worked example for a sample of data underlying this report:Step 2:calculate intersectio

278、n points between AI capabilities and as-is work profile of the sector1 year5 yearsControls Testing10%60%Policy Enforcement20%55%Drafting Legal Documents15%55%Regulatory Thematic Reviews20%50%Regulatory Compliance10%40%Licensing and Registration10%40%Legal&ComplianceTotal AI impact by task46Financial

279、&Professional Services:The Future of AI&the WorkforceData ProcessingSummarising InformationCreating ContentDiary MngtChatbotsCodeGenTranslationControls Testing17%17%9%17%Policy Enforcement17%17%13%8%Drafting Legal Documents15%26%7%7%Regulatory Thematic Reviews7%21%21%Regulatory Compliance10%20%10%Li

280、censing and Registration7%13%13%7%AI Impact by AI capability(sums to 5 year AI impact)This is analysis was run for 471 representative tasks across the FPS sector.47Financial&Professional Services:The Future of AI&the WorkforceSpecial thanks Special thanks to the industry experts who gave their time

281、for the report alongside the following KPMG subject matter experts who contributed to our report.City of London Corporation Marco CiaoKerstin MathiasKPMG UKLeanne AllenDavid CummingsDouglas DickHuw EvansJohn HallsworthPaul HenningerDavid RowlandsPaul GreenanOlivia LarkinHead of Skills and Workforce

282、PolicyThe City of London CorporationOlivia.Larkincityoflondon.gov.ukErik DronenSenior Regulation and Business Enviroment AdviserThe City of London CorporationErik.Dronencityoflondon.gov.ukMel NewtonPeople Consulting PartnerKPMG UKMel.Newtonkpmg.co.uk Matt YerburyPeople Consulting KPMG UKMatthew.Yerb

283、urykpmg.co.ukJackie ToddPeople ConsultingKPMG UKJackie.ToddKPMG.co.ukNiale CleoburyPeople Consulting KPMG UKNiale.Cleoburykpmg.co.uk David McIntoshPeople ConsultingKPMG UKDavid.McIntoshKPMG.co.uk ContactsFor any press or media enquiries,please contact Gerard Swinley gerard.swinleykpmg.co.uk(KPMG)orT

284、homas.Kingsleycityoflondon.gov.uk(CoLC).48Financial&Professional Services:The Future of AI&the WorkforceReferences1.Vision for Economic Growth(2023)https:/www.theglobalcity.uk/PositiveWebsite/media/Research-reports/Vision_for_Economic_Growth.pdf2.Our global offer to business(2024)https:/www.thegloba

285、lcity.uk/PositiveWebsite/media/Research-reports/Our-global-offer-to-business-2024.pdf 3.AI-Accelerating Innovation-City of London(2024)https:/www.cityoflondon.gov.uk/supporting-businesses/economic-research/research-publications/ai-accelerating-innovation4.FSSC Future Skills Report(2024)(financialser

286、vicesskills.org)5.GOV.UK Impact of AI on UK jobs and training(2023)https:/assets.publishing.service.gov.uk/media/656856b8cc1ec500138eef49/Gov.UK_Impact_of_AI_on_UK_Jobs_and_Training.pdf6.ONS,2023;Vacancies and Jobs in the UK(2023)https:/www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmen

287、tandemployeetypes/bulletins/jobsandvacanciesintheuk/october2023 7.FSSC Member Survey(2022)https:/financialservicesskills.org/research/8.Microsoft Workplace Trends Index,(2023)https:/ Workplace Trends Index(2023)https:/ Leave Report(2024)https:/ of Jobs(2023)https:/www.weforum.org/publications/the-fu

288、ture-of-jobs-report-2023/12.Talent Insight Group(2022)The Cost of Recruitment https:/wp.financialservicesskills.org/wp-content/uploads/2022/01/Reskilling-A-business-case-FINAL-EMBARGOED.pdf13.UK Digital Strategy(2022)https:/www.gov.uk/government/publications/uks-digital-strategy14.OECD Artificial In

289、telligence(2023)Job Quality and Inclusiveness https:/oecd.org/employment-outlook/2023/Research cited throughout the report is document below:49Financial&Professional Services:The Future of AI&the WorkforceReferencesIn addition,the following sources formed part of a wider literature review:KPMG,The f

290、uture of AI in finance(2023)https:/ Intelligence in Risk Management(2021)https:/ generative AI spark innovation in asset and wealth management(2023)https:/ of AI ups the ante on governance(2023)https:/ KPMG,Trusted AI and the Regulatory Landscape(2024)https:/ called to Duty in the GI sector,(2024)ht

291、tps:/ insurance companies work with IBM to implement generative AI-based solutions(2024)https:/ Reuters,Legal tech AI:Generative AIs impact on legal departments(2023)https:/ Areas Of Impact For AI In Fintech(2021)https:/ Future Of AI In Banking(2023)https:/ AI transforming wealth and asset managemen

292、t(2023)https:/ of Artificial Intelligence(AI)in Investment Management(2023)https:/ technologies shaping the future of fintech(2021)https:/ impact of Artificial Intelligence on portfolio management(2023)https:/ Law Experts,Will Artificial Intelligence(AI)Displace Legal Practitioners(2023)https:/ AI T

293、o Bring Fairness To Mortgage Lending(2021)https:/ sites/donnafuscaldo/2019/03/19/zestfinance-using-ai-to-bring-fairness-to-mortgage-lending/?sh=192901367f2dBinariks,AI in Fintech:Top 8 Use Cases with Examples(2023)https:/ News,How AI is affecting Fintech(2022)https:/www.fintechnews.org/how-ai-is-aff

294、ecting-fintech/50Financial&Professional Services:The Future of AI&the WorkforceReferencesIfford,How Generative AI is Enhancing Risk Management&Fraud Detection(2023)https:/ Claims Process is Changing due to GenAI(2023)https:/ banks can harness the power of GenAI-Explosive Growth Anticipated for the G

295、enerative AI Market with a Remarkable CAGR of 31.3%by 2033 (2023)https:/ Global,AI in Banking:AI Will Be An Incremental Game Changer(2023)https:/ Generative AI In Wealth And Asset Management(2023)https:/ Infra,Artificial Intelligence in Professional Services Market Infra Consulting https:/marketinfr

296、a.co.uk/ai-in-professional-servicesForbes,Harnessing The Power Of AI In The Insurance Sector(2023)https:/ Journal of Law Management and Humanities,Impact of Artificial Intelligence on Legal Industry-International Journal of Law Management&Humanities(2024)https:/ Reuters,AI will transform every aspec

297、t of professional work,new“Future of Professionals”report shows(2023)https:/ insurers can leverage the power of generative AI(2023)https:/ Intelligence and Machine Learning in Asset Management(2019) Times,AI in law,consulting and professional services(2023)https:/ 51Financial&Professional Services:T

298、he Future of AI&the WorkforceDisclaimerThe information contained herein is of a general nature and is not intended to address the circumstances of any particular individual or entity.Although we endeavour to provide accurate and timely information,there can be no guarantee that such information is a

299、ccurate as of the date it is received or that it will continue to be accurate in the future.No one should act on such information without appropriate professional advice after a thorough examination of the particular situation.2024 KPMG LLP,a UK limited liability partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.The KPMG name and logo are trademarks used under license by the independent member firms of the KPMG global organisation.